Context-aware energy neutral AoI scheduling for IoT over LEO satellite networks.

Abstract

Low Earth orbit (LEO) non-terrestrial networks (NTNs) enable wide-area status monitoring for energy harvesting Internet of Things (EH-IoT) devices, but maintaining fresh information is challenging under intermittent satellite access, shared uplink airtime, and time-varying energy availability. We formulate joint status-update scheduling and node–satellite association as a long-term Age of Information (AoI) minimization problem subject to update availability, satellite visibility, residual contact-time feasibility, battery energy causality, shared satellite airtime, and long-term energy neutrality constraints. To address this time-coupled problem online, we propose Context-Aware Lyapunov Energy-Neutral Scheduling (CALENS), a training-free framework combining Lyapunov energy-deficit control with freshness urgency, predictable contact gaps, correlation- and recency-aware update informativeness, and battery-adaptive prioritization. A persatellite knapsack procedure converts the resulting utilities into feasible node–satellite assignments. Trace-driven simulations show that CALENS improves average and effective AoI, reduces redundant transmissions, and operates closer to finite-horizon energy balance than the evaluated baselines, without requiring offline policy training.